Anomalous Diffuse CO2 Emission prior to the January 2002

Short-term Unrest at San Miguel Volcano, El Salvador, Central

America

NEMESIO M. PEREZ,1 PEDRO A. HERNANDEZ,1 ELEAZAR PADRON,1

RAFAEL CARTAGENA,2 RODOLFO OLMOS,2 FRANCISCO BARAHONA,2 GLADYS MELIAN,1

PEDRO SALAZAR,1 and DINA L. LOPEZ3

Abstract—On January 16, 2002, short-term unrest occurred at San Miguel volcano. A gas-and-steam-

ash plume rose a few hundred meters above the summit crater. An anomalous microseismicity pattern,

about 75 events between 7:30 and 10:30 hours, was also observed. Continuous monitoring of CO2 efflux on

the volcano started on November 24, 2001, in the attempt to provide a multidisciplinary approach for its

volcanic surveillance. The background mean of the diffuse CO2 emission is about 16 g m)2 d)1, but a 17-

fold increase, up to 270 g m)2 d)1, was detected on January 7, nine days before the January 2002 short-

term unrest at San Miguel volcano. These observed anomalous changes on diffuse CO2 degassing could be

related to either a sharp increase of CO2 pressure within the volcanic-hydrothermal system or degassing

from an uprising fresh gas-rich magma within the shallow plumbing system of the volcano since

meteorological fluctuations cannot explain this observed increase of diffuse CO2 emission.

Key words: San Miguel, volcanic activity, diffuse degassing, carbon dioxide.

Introduction

The symmetrical cone of San Miguel, one of the most active volcanoes in El

Salvador (Central America) rises from near sea level to 2,132 meters of elevation. This

basaltic volcano is located in the eastern part of the country and lies on the southern

fault of the Central American graben at the intersection with NW-SE trending faults

(Fig. 1). San Miguel volcano has erupted at least 30 times since 1699, all events

classified with aVEI of 1 or 2.Many of the earlier eruptions occurred at flank vents, but

since 1867 all have taken place at the summit (MEYER-ABICH, 1956; SIMKIN and

SIEBERT, 1994). San Miguel volcano has had active fumaroles in the summit region at

1Environmental Research Division, Instituto Tecnologico y de Energıas Renovables (ITER), 38611,Granadilla, S/C de Tenerife, Spain

2Instituto de Ciencias de la Tierra, Universidad de El Salvador, El Salvador, Central America3Department of Geological Sciences, 316 Clippinger Laboratories Ohio University, Athens, OH,

45701, USA

Pure appl. geophys. 163 (2006) 883–8960033–4553/06/040883–14DOI 10.1007/s00024-006-0050-1

� Birkhauser Verlag, Basel, 2006

Pure and Applied Geophysics

least since 1964, and their variable intensity sustains usually a faint plume emitted from

its summit crater. Since 1970, weak explosions took place four times, the last in January

2002 (GVN BULLETIN, 2002).

The January 2002 short-term volcanic unrest at San Miguel was characterized by

the emission of a gas-and-steam plume of 100 metric tons/day of sulfur dioxide

containing a slight amount of ash rising with a mushroom-like profile a few hundred

meters above the summit crater of San Miguel (GVN BULLETIN, 2002). In addition,

the event was accompanied by anomalous microseismicity with about 75 events

between 7:30 and 10:30 hours, on January 16, 2002. This seismic swarm was followed

by progressively decreasing seismic activity during the next days. This style of

increased seismicity and gas emission is within the range of normal activity at San

Miguel according to SNET (Servicio Nacional de Estudios Territoriales; the

National organization in-charge of the volcano monitoring at El Salvador) since

intermittent periods of vigorous steam-and-gas emission from San Miguel have been

commonly reported in recent years (GVN BULLETIN, 2002). The population at risk

from a San Miguel eruption with significant ashfall is a mix of urban and rural

residents. The city of San Miguel, at the foot of the NE flank of the volcano, has a

population of �150,000, and the rural zone that would likely be affected has a

population of �100,000 (GVN BULLETIN, 2002).

Figure 1

San Miguel volcano located in the eastern part of El Salvador, Central America. Closed triangules

represent those active volcanic systems with a continuous geochemical monitoring station of CO2 efflux.

884 N.M. Perez et al. Pure appl. geophys.,

Scientists have long recognized that gases dissolved in magma provide the driving

force of volcanic eruptions, but only recently geochemical techniques permitted

continuous measurement of different types and manifestations of volcanic gases

released into the atmosphere. Carbon dioxide is the major gas species after water

vapor in both volcanic fluids and magmas and it is an effective tracer of subsurface

magma-degassing, due to its low solubility in silicate melts (GERLACH and GRAEBER,

1985). Since this degassed CO2 travels upward by advective-diffusive transport

mechanisms and manifests itself at the ground surface, soil CO2 efflux pattern

changes over time provide information about subsurface magma movement

(HERNANDEZ et al., 2001a; CARAPEZZA et al., 2004).

Figure 2

Simple morphological and structural map of San Miguel volcano and its surroundings. This map shows

major inhabited areas in the well as the location of the Las Placitas’ fumarole and the geochemical station

(closed circle). Solid lines represent faults and dashed lines represent inferred fault systems. Dotted lines

represent major highways.

Vol. 163, 2006 Anomalous Diffuse CO2 Emission 885

Extensive work on diffuse CO2 degassing has been performed at volcanic and

geothermal areas over the last 13 years, suggesting that even during periods of

quiescence, volcanoes release large amounts of carbon dioxide in a diffuse form

(BAUBRON et al., 1990; ALLARD et al., 1991; FARRAR et al., 1995; CARTAGENA et al.,

2004; CHIODINI et al., 1996, 1998; HERNANDEZ et al., 1998, 2001b; GERLACH et al.,

1998; SOREY et al., 1998; GIAMMANCO et al., 1998; PEREZ et al., 1996, 2004; ROGIE

et al., 2001; SALAZAR et al., 2001). However, few works related to continuous

monitoring of the diffuse CO2 degassing have been published (ROGIE et al., 2001;

SALAZAR et al.,2002, 2004; MORI et al., 2002; GRANIERI et al., 2003; CARAPEZZA

et al., 2004; BRUSCA et al., 2004).

Conventional geophysical methods are operating for the volcano monitoring

program of SanMiguel (e.g., volcano seismicity). In order to provide a multidisciplin-

ary approach to volcanic surveillance of San Miguel, the Spanish Aid-Agency (AECI)

provided the financial-aid to set up a geochemical station for the continuous

monitoring of diffuse CO2 degassing. Similar monitoring stations were installed at

SantaAna-Izalco-Coatepeque, SanSalvador andSanVicente volcanic systems in 2001.

Procedures and Methods

With the aim of detecting anomalous temporal variations in diffuse CO2 emission

rates related to changes of volcanic activity at San Miguel, an automatic geochemical

station (WEST Systems, Italy) was installed on November 24, 2001. This station

continuously monitors diffuse CO2 degassing on the eastern flank of San Miguel

volcano (594 m a.s.l., Fig. 2). The station is equipped with an on-board microcom-

puter as well as sensors to measure the air CO2 gas concentration, wind speed and

direction, air temperature and relative humidity (1 m above the ground). The

accumulation chamber is lowered on to the ground for several minutes every hour,

and during this period the gas is continuously extracted from the chamber, sent to the

IR spectrophotometer, and then injected again into the chamber. The latter is

equipped with a mixing device in order to improve gas mixing. Soil CO2 efflux is then

measured according to the accumulation chamber method by means of a NDIR

(non-dispersive infrared) spectrophotometer (Drager Polytron IR transmitter) with a

double-beam IR detector with solid state sensor compensated in temperature.

Accuracy of 3% is acquired for a reading at 350 ppm. The automatic geochemical

station is powered with a solar cell panel and a backup battery. Values of CO2 efflux

(g m)2 d)1) are estimated from the rate of concentration increase in the chamber at

the observation site, accounting for changes of atmospheric pressure and temper-

ature to convert volumetric concentrations to mass concentrations. The chamber rim

is designed to be set properly on the ground in order to eliminate the input of

atmospheric air, which could cause significant errors, especially on windy days. The

reproducibility for the range 10–20,000 g m)2 d)1 is ±10%. We assumed a random

886 N.M. Perez et al. Pure appl. geophys.,

error of ±10% in the emission rates of CO2, based on variability of the replicate

measurements carried out on known CO2 efflux rates in the laboratory. All

measurements were performed during the dry season in El Salvador.

Results and Discussion

A time series of 2,326 hourly observations of diffuse CO2 emission rate, wind

speed and barometric pressure from 17:00 hours of November 24, 2001, to

14:00 hours of March 1, 2002, as well as their moving averages of 36 hours are

shown in Figure 3. A total of 12.4% of missing data throughout the observation

Figure 3

Time series of (a) diffuse CO2 emission rate, (b) wind speed and (c) barometric pressure on the eastern flank

of San Miguel Volcano from November, 2001 to March, 2002.

Vol. 163, 2006 Anomalous Diffuse CO2 Emission 887

period occurred due to telemetry problems. The effect of the missing data on the

overall data set is not likely to produce spurious correlation results because most of

the missing data were recorded when diffuse CO2 degassing on the eastern flank of

San Miguel volcano was stationary. Table 1 summarizes the result of the CO2 efflux

data and the observed external variables.

The observed diffuse CO2 emission data ranged from negligible values up to 270 g

m)2 d)1. These time series may be separated into four distinct set intervals: (i) from

the beginning of the observation until January 6, 2002, characterized by a low

variance of the CO2 efflux and relatively low diffuse CO2 emission rates of about

16 g m)2, d)1; (ii) from January 6 to 16, 2002, characterized by a short-term transient

in the soil CO2 degassing emission � 9 days before the short-term unrest with CO2

efflux values reaching up to 270 g m)2 d)1; (iii) from January 16 to February 9, 2002,

characterized by wider scatter of efflux data and a relatively high CO2 efflux value

(143 g m)2 d)1) occurring at the onset of an anomalous seismic pattern at San

Miguel volcano in January 16, 2002, and (iv) a period from February 9 to March 1,

2002, characterized by CO2 efflux values similar to those observed within period (i).

Wind speed values ranged from 0 to 3.8 m/s. Average wind speed was 0.18 m/s.

However, wind speed time series showed a simultaneous increase with the transient

of soil CO2 efflux in period (ii). Barometric pressure ranged from 938 to 949 HPa

within periods (i) to (iii). Barometric pressure time series reflected the passing of low

Table 1

Summary of results from the geochemical station at San Miguel Volcano

CO2 efflux (g m)2 d)1) Range <0.25)270Mean 29.4

Std deviation 33.2

Air relative humidity (%) Range 12.5–97.3

Mean 54.2

Std deviation 19.3

Air temperature (�C) Range 15.9–34.5

Mean 23.9

Std deviation 4.2

Bar pressure (hPa) Range 932.1–959.0

Mean 945.3

Std deviation 2.7

Wind speed (m s)1) Range 0.0–13.7

Mean 1.1

Std deviation 1.6

Wind direction (�) Range 0.8–341.7

Mean 202.1

Std deviation 90.5

Tidal ( ) Range (x1000) )999.3–2241.7Mean (x 1000) 189.7

Std dev (x1000) 754.2

888 N.M. Perez et al. Pure appl. geophys.,

and high pressure fronts over the observation site. A high pressure atmospheric

regime started on January 2 and ended January 12, 2002. Both wind speed and

barometric pressure displayed high values during period (ii), where a short-term

transient in the soil CO2 degassing was observed. These observations drive the

question whether wind speed and/or barometric pressure may be responsible for the

observed changes in the soil CO2 degassing at San Miguel volcano. At this aim, time

domain classical techniques were applied to get the autocorrelation functions of the

three original time series, showing the presence of cycles of the same length.

Because various types of serial dependence (trends, cycles, and autoregressive

serial dependence) within time series can give rise to spurious or misleading

correlations between time series (WARNER, 1998), soil CO2 efflux, wind speed and

barometric pressure time series were standardized and prewhitened before spectral

analysis was performed. Soil CO2 efflux time series was log-transformed and

differenced one time to obtain a much more stationary time series. Wind speed and

barometric pressure time series only required one order differencing before getting

stationarity. Then spectral coherences between the prewhitened soil CO2 efflux and

wind speed, as well as soil CO2 efflux and barometric pressure time series were

computed (Fig. 4). Cross-spectral analysis of the prewhitened time series shows that

a high percentage of shared variance between both sets of time series occur at one

and two cycles per day, which correspond’s to periods of 24 and 12 hours,

respectively, suggesting a causal influence or coupling of wind speed and barometric

pressure on the observed soil CO2 efflux values at diurnal and semi-diurnal

frequencies. These results suggest a close relationship between the temporal

variations of the soil CO2 efflux and those of meteorological variables, showing

Figure 4

Cross-spectral analysis of the soil CO2 efflux and wind speed (solid line) and soil CO2 efflux and barometric

pressure (dashed line) time series.

Vol. 163, 2006 Anomalous Diffuse CO2 Emission 889

24- and 12-hour periodicities (i.e., wind speed, barometric pressure, solar radiation,

air temperature, etc.).

Once the main cyclic components of the prewhitened soil CO2 efflux time series

have been identified, it was possible to model the original soil CO2 efflux signal using

the harmonic analysis (Fig. 5a), based on the following equation

UðtÞ ¼ lþX2

i¼1Ai cos -itð Þ þ Bi sin -itð Þf g þ e;

where F(t) and l are the observed soil CO2 efflux time series and its mean,

respectively, xi (i = 1,2) are the main identified frequencies corresponding to the

main periodic components of the soil CO2 efflux signal, Ai and Bi are model

Figure 5

(a) Original soil CO2 efflux signal (closed circles) and fitted harmonic model (solid line) with two major

periodic components; (b) residual of harmonic analysis showing the persistence of a large amplitude

transient in the soil CO2 time series.

890 N.M. Perez et al. Pure appl. geophys.,

parameters to adjust during the fitting process, t is the observation time and, finally, �

are the residuals uncorrelated with the cos and sin terms. Using this simple

deterministic model, the autoregressive behavior of the soil CO2 efflux time series,

which is linked to the periodic behavior of the meteorological variables (wind speed

and barometric pressure), can be removed from the original soil CO2 efflux signal. As

a result of this filtering procedure, a large amplitude transient in the soil CO2

degassing still remains within the residuals (Fig. 5b) temporally related with the onset

of the anomalous seismic pattern at San Miguel volcano.

Low amplitude and almost universal short-term CO2 efflux changes driven by

meteorological fluctuations have been observed in others volcanic systems (SALAZAR

et al., 2000, 2002; PADRON et al., 2001; ROGIE et al., 2001; GRANIERI et al., 2003).

This significant transient increase of CO2 efflux rate observed on January 6 and 7,

2002, cannot be explained in terms of meteorological fluctuations such as barometric

pressure and wind speed temporal variations recorded at the station site.

Since meteorological fluctuations cannot explain the observed relative increase of

the CO2 efflux rate on January 6 and 7, 2002 (Fig. 3), it is evident that the observed

changes of diffuse CO2 emission rate on the eastern flank of San Miguel volcano

should be related to either a sharp increase of CO2 pressure within the volcanic-

hydrothermal system or degassing from an uprising fresh gas-rich magma within the

shallow plumbing system of the volcano as it was explained for the observed changes

on CO efflux prior to the Stromboli 2002–2003 eruptive events (CARAPEZZA et al.,

2004; BRUSCA et al., 2004). As an alternative procedure to isolate the joint impulse-

response function of the CO2 efflux and determine the effects of variations in the

hydrothermal system on diffuse CO2 efflux, we used multivariate regression analysis

(MRA) to delineate the relations between CO2 efflux and external factors and then

use these relations to filter out the effects of these factors on the measured efflux

history. In this case, we considered in the analysis the most relevant parameters to

explain the variability of the CO2 efflux, barometric pressure, wind speed, tidal, air

temperature, air relative humidity, wind direction and power supply. MRA is

commonly used to predict a dependent variable, y, as a function of relevant

Table 2

Stepwise forward results From the MRA analysis

Multiple R F to ENTER p-level

Air temperature 0.132472 36.331 0.0000

TIDAL 0.178912 30.371 0.0000

Barometric pressure 0.191681 9.981 0.0016

Wind Speed 0.199183 6.201 0.0128

Power supply 0.200368 1.001 0.3170

Wind direction 0.200887 0.440 0.5071

Relative humidity 0.201147 0.221 0.6381

Vol. 163, 2006 Anomalous Diffuse CO2 Emission 891

Figure 6

Observed (a) and predicted (c) CO2 efflux time series of San Miguel Volcano. Figures 6d and 6e show the

total residuals and smoothed residual of the MRA for the same data series.

892 N.M. Perez et al. Pure appl. geophys.,

explanatory variables, x1, x2,…, xn. Results of MRA provide an understanding of

the percentage of the variability in y that is explained by the selected set of x

variables. A forward stepwise regression analysis (Table 2) was performed to identify

the most efficient set of x variables that would explain most of the variability in y and

also to assess whether the relationship between y and the selected set of x variables is

statistically significant, with the retained explanatory variables. A F to enter criterion

of 1.0 and F to remove 0.0 for entering or removing an explanatory variable were

used. The correlation coefficient (r) between the CO2 efflux and each explanatory

variable was calculated. The highest simple correlation was between the CO2 efflux

and air temperature (r = 0.13). The statistical significance level of the regression

model (p value) was less than 0.0001 with air temperature and TIDAL as

independent variables, indicating a highly significant model. The square of the

multiple regression coefficients (r2) for the regression equation was 0.20, which means

that about 20% of the variability in the estimated CO2 efflux is explained by the

regression model. The time evolution of the residuals of MRA filtering is shown in

Figure 6. Again, the filtered trend and residual components showed the anomalous

increase on the diffuse CO2 degassing 10 days before the onset of the short-term

unrest at San Miguel volcano.

Fluid pressure fluctuations in volcanic systems might be also related to stress/

strain changes in the subsurface due to either magma rising beneath the volcano or

the occurrence of a relatively high magnitude earthquake in the vicinity of the

volcano. The last rationale was recently described by SALAZAR et al. (2002) at San

Vicente volcano (El Salvador, Central America) where significant CO2 efflux rate

fluctuations were measured relative to a 5.1 magnitude earthquake, which occurred

to 25 km away from the observation site on May 8, 2001. A significant increase in

CO2 efflux rate was mainly driven by strain changes prior to this earthquake.

Since the seismic activity recorded in the vicinity of San Miguel volcano at the

end of 2001 and early 2002 do not reflect the occurrence of a relatively high

magnitude earthquake, magma movement at depth seems the most reasonable

scenario for the observed increase on the diffuse CO2 emission rate at San Miguel

volcano on January 6 and 7, 2002. This geological phenomena should be responsible

for a relative increment of the internal pressure of magmatic volatiles due to magma

rising within the volcanic-hydrothermal system; therefore, driving an increase of

diffuse CO2 emission in the surface environment of San Miguel volcano prior to the

observed vigorous steam-and-gas emission through the plume and microseismicity

changes on January 16, 2002.

After the observed precursory geochemical signature of diffuse CO2 emission rate

at San Miguel volcano, period (iii), CO2 efflux values showed a higher dispersion

than those observed before January 6, 2002. This scattered behavior on the CO2

efflux rate could mainly be due to long-period earthquakes, volcanic tremor, and

explosion events which were recorded at San Miguel in late January and February

(GVN BULLETIN, 2002).

Vol. 163, 2006 Anomalous Diffuse CO2 Emission 893

Conclusions

An increase up to 17-fold the background mean value of diffuse CO2 emission

rate was observed before short-term unrest at San Miguel volcano on January 16,

2002. These CO2 efflux changes were not driven by fluctuations of meteorological

variables such as wind speed and barometric pressure and seem clearly associated to

a relative increment of CO2 pressure within the volcanic-hydrothermal system due to

magma movement at depth beneath San Miguel volcano. These results showed the

potential of applying continuous monitoring of soil CO2 efflux to improve and

optimize the detection of early warning signals of future volcanic crisis at San Miguel

as well as in other active volcanic systems. Further observations, after fixing the

radiotelemetry system are needed to verify the existence of a close relationship

between diffuse CO2 emission rate and volcanic activity.

Acknowledgements

The authors are grateful to the Office of the Spanish Agency for International

Cooperation (AECI) in El Salvador, the Spanish Embassy in El Salvador, The

Ministry of Environment and Natural Resources of the Government of El Salvador,

and Geotermica Salvadorena (GESAL) for their assistance during the fieldwork. We

thank Jose M. Salazar for constructive suggestions to improve the paper and Tomas

Soriano for help in the field work. We are indebted to the GRP of El Salvador’s Civil

National Police for providing security during our stay in El Salvador. This research

was mainly supported by the Spanish Aid Agency (Agencia Espanola de Cooper-

acion Internacional – AECI), and additional financial-aid was provided by the

Cabildo Insular de Tenerife, CajaCanarias (Canary Islands, Spain), the European

Union, and The University of El Salvador.

REFERENCES

ALLARD, P., CARBONELLE, J., DAJLEVIC, D., LE BRONCE, J., MOREL, P., ROBE, M.C., MAURENADS, J.M.,

FAIVRE-PIERRET, R., MARTIN, D., SABROUX, J.C., and ZETTWOOG, P. (1991), Eruptive and diffuse

emissions of CO2 from Mount Etna, Nature 351, 387–391.

BAUBRON, J.C., ALLARD, P., and TOUTAIN, J.P. (1990), diffuse volcanic emissions of carbon dioxide from

Vulcano Island, Italy, Nature 344, 51–53.

BRUSCA, L., INGUAGGIATO, S., LONGO, M., MADONIA, P., and MAUGERI R. (2004), The 2002–2003

eruption of Stromboly (Italy): Evaluation of the volcanic activity by means of continous monitoring of soil

temperature, CO2 flux, and meteorological parameters, Geochem., Geophys. Geosyst. 5, Ql2001,

doi:10.1029/2004GC000732.

894 N.M. Perez et al. Pure appl. geophys.,

CARAPEZZA, M. L., INGUAGGIATO, S., BRUSCA, L., and LONGO, M. (2004), Geochemical precursors of the

activity of an open-conduit volcano: The Stromboli 2002–2003 eruptive events, Geophys. Res. Lett. 31,

L07620, doi:10.1029/2004GL019614.

CARTAGENA, R., OLMOS, R., LoPEZ, D., BARAHONA, F., SORIANO, T., HERNANDEZ, P.A., and PEREZ, N.M.

(2004), Diffuse degassing of carbon dioxide, radon and mercury at San Miguel Volcano, El Salvador,

Central America, Bull. Geol. Soc. America 375, 203–212.

CHIODINI, G., FRONDINI, F., and RACO, B. (1996), Diffuse emission of CO2 from the Fossa Crater, Vulcano

Island (Italy), Bull. Volcanol. 58, 41–50.

CHIODINI, G., CIONI, R., GUIDI, M., RACO, B., and MARINI, L. (1998), Soil CO2 flux measurements in

volcanic and geothermal areas, Appl. Geochem. 13, 543–552.

FARRAR, C.D., SOREY, M.L., EVANS, W.C., HOWLE, J.F., KERR, B.D., KENNEDY, B.M., KING, Y., and

SOUTHON, J.R. (1995), Forest-killing diffuse CO2 emission at Mmammoth Mountain as a sing of magmatic

unrest, Nature 376, 675–678.

GERLACH, T.M.J. and GRAEBER, E.J. (1985), Volatile budget of Kilauea Volcano, Nature 313, 273–277.

GERLACH, T.M.J., DOUKAS, M.K., MCGEE and KESSLER, R. (1998), Three-year decline of magmatic CO2

emission from soils of a Mammoth Mountain tree kill: Horseshoe Lake, CA, 1995–1997, Geophys. Res.

Lett. 25, 1947–1950.

GIAMMANCO, S., GURRIERI, S., and VALENZA, M. (1998), Anomalous soil CO2 degassing in relation to faults

and eruptive fissures on Mount Etna (Sicily), Bull. Volcanol. 69, 252–259.

GRANIERI, D., CHIODINI, G., MARZOCCHI, W., and AVINO, R. (2003). Continuous monitoring of CO2 soil

diffuse degassing at Phlegraean Fields (Italy): Influence of environmental and volcanic parameters, Earth

Planet. Sci. Lett. 212, 167–179.

GVN BULLETIN (2002), Minor gas-and-ash emission in January 2002; Summary of earlier activity, Bull.

Global Volcan. Network 27, 02.

HERNANDEZ, P.A., PEREZ, N.M., SALAZAR, J.M., NOTSU, K., and WAKITA, H. (1998), Diffuse emission of

carbon dioxide, methane, and helium-3 from Teide Volcano, Tenerife, Canary Islands, Geophys. Res. Lett.

25, 3,311–3,314.

HERNANDEZ, P.A., SALAZAR, J.M., SHIMOIKE, Y., MORI, T., NOTSU, K., PEREZ, N.M. (2001b), Diffuse

emission of CO2 from Miyakejima Volcano, Japan, Chem. Geology 177, 175–185.

HERNANDEZ, P.A., NOTSU, K., SALAZAR, J.M., MORI, T., NATALE, G., OKADA, H., VIRGILI, G., SHIMOIKE,

Y., SATO, M., and PEREZ, N.M. (2001a), Carbon dioxide degassing by advective flow from Usu Volcano,

Japan, Science 292, 83–86.

ISHIGURO, M., SATO, T., TAMURA, Y., and OOE, M. (1984), Tidal data analysis: An introduction to Bbaytap,

Proc. Institute of Statist. Math. 32, 71–85.

MEYER-ABICH, H. (1956), Los Volcanes Activos De Guatemala Y El Salvador, (America Central), Anales

Del Servicio Geologico Nacional De El Salvador, pp. 3–129, Ministerio De Obras Publicas. Republica

De El Salvador.

MORI, T., NOTSU, K., HERNANDEZ, P.A., SALAZAR J, M.L., PEREZ, N.M., VIRGILI, G., SHIMOIKE, Y., and

OKADA, H. (2002), Continuos monitoring of soil CO2 efflux from the summit region of Usu Volcano, Japan,

Bull. Volcanol. Soc. Japan 47, 339–345.

PADRON, E., SALAZAR, J.M.L., HERNANDEZ, P.A., and PEREZ, N.M. (2001), Continuous monitoring of

diffuse CO2 degassing from Cumbre Vieja Volcano, La Palma, Canary Islands, EOS, Trans. Am.

Geophys. Union 82, F1132.

PEREZ, N.M., WAKITA, H., LOLOK, D., PATIA, H., TALAI, B., and MCKEE, C.O. (1996), Anomalous soil gas

CO2 concentrations and relation to seismic activity at Rabaul Caldera, Papua New Guinea, Geogaceta 20,

1000–1003.

PEREZ, N.M., SALAZAR, J.M.L., HERNANDEZ, P.A., SORIANO T., LOPEZ, K., and NOTSU, K. (2004), Diffuse

CO2 and222rn dcegassing from San Salvador Volcano, El Salvador, Central America, Bull. Geol. Soc. Am.

375, 227–236.

ROGIE, J.D., KERRICK, D.M., SOREY, M.L., CHIODINI, and GALLOWAY, D.L. (2001), Dynamics of carbon

dioxide emission at Mammoth Mountain, California, Earth Planet. Sci. Lett. 188, 535–541.

SALAZAR, J.M.L., PEREZ, N.M., and HERNANDEZ, P.A., (2000), Secular variations of soil CO2 flux levels at

the summit cone of Teide Volcano, Tenerife, Canary Islands, EOS, Trans. Am. Geophys. Union 81,

F1317.

Vol. 163, 2006 Anomalous Diffuse CO2 Emission 895

SALAZAR, J.M.L., HERNANDEZ, P.A., PEREZ, N.M., MELIAN, G., ALVAREZ, J., SEGURA, F., and NOTSU, K.

(2001), Diffuse emission of carbon dioxide from Cerro Negro Volcano, Nicaragua, Geophys. Res. Lett. 28,

4275–4278.

SALAZAR, J.M.L., PEREZ, N.M., HERNANDEZ, P.A., SORIANO, T., BARAHONA, F., OLMOS, R., CARTAGENA,

R., LoPEZ, D., LIMA, N., MELIAN, G., GALINDO, I., PADRON, E., SUMINO, H. and NOTSU, K., (2002),

Precursory diffuse carbon dioxide degassing signature related A 5.1 magnitude earthquake in El Salvador,

Central America, Earth Planet. Sci. Lett. 205, 81–89.

SALAZAR, J.M.L., HERNANDEZ, P.A., PEREZ, N.M., OLMOS, R., BARAHONA, F., CARTAGENA, R.,

SORIANO, T., LOPEZ, K., and NOTSU, K. (2004), Spatial and temporal variations of diffuse CO2

degassing at Santa Ana-Izalco-Coatepeque Volcanic Complex, El Salvador, Central America, Bull.

Geolog. Soc. Am. 375, 135–146.

SIMKIN, T., and SIEBERT, L., Volcanoes of The World, 2nd Ed. (Geosci. Press. Tuscon, Arizona 1994), 349

pp.

SOREY, M.L., EVANS, W.C., KENNEDY, B.M., FARRAR, C.D., HAINSWORTH, L.J., and HAUSBACK, B.

(1998), Carbon dioxide and helium emissions from a reservoir of magmatic gas beneath Mammoth

Mountain, California, J. Geophys. Res. 103(B7), 15,303–15,323.

WARNER, REBECCA, M. Spectral analysis of time-series data (Guilford Press 1998), 225 pp.

(Received: December 1, 2004; revised: November 18, 2005; accepted: November 28, 2005)

To access this journal online:

http://www.birkhauser.ch

896 N.M. Perez et al. Pure appl. geophys.,